meaning-memory 0.1.0

Significance-weighted memory for AI agents

meaning-memory

Significance-weighted memory for AI agents.

pip install meaning-memory

Quick Start

from pathlib import Path
from meaning_memory import LedgerEntry, FileBackend, load_config

# Load config from TOML
config = load_config("my-agent", config_dir=Path("config/"))

# Create backend
backend = FileBackend.from_rosebud_config(config)

# Append a memory
entry = LedgerEntry(
    date="2026-04-07", category="fact",
    text="The sky is blue",
    mem_id="a1b2c3d4e5f67890", source="manual",
    status="active", sig=0.70, pinned=False, raw_line="",
)
backend.append_memory(entry)

# Query active memories
for entry in backend.get_active_entries():
    print(f"[{entry.sig:.2f}] {entry.text}")

CLI

meaning-memory config validate my-agent --config-dir ./config
meaning-memory config inspect my-agent --config-dir ./config
meaning-memory config generate my-agent --config-dir ./config
meaning-memory version

Requirements

• Python 3.11+ (stdlib tomllib)
• Zero external dependencies

License

Apache-2.0 — see LICENSE file.

---

Built by StarkMind — Human-AI Collaboration Lab